The service quality of a passenger transport operator can be measured through face-to-face surveys at the terminals or on board. However, the resulting responses may suffer from the influence of the intrinsic aspects of the respondent's personality and emotional context at the time of the interview. This study proposes a methodology to generate and select control variables for these latent psychosituational traits, thus mitigating the risk of omitted variable bias. We developed an econometric model of the determinants of passenger satisfaction in a survey conducted at the largest airport in Latin America, S\~ao Paulo GRU Airport. Our focus was on the role of flight delays in the perception of quality. The results of this study confirm the existence of a relationship between flight delays and the global satisfaction of passengers with airports. In addition, favorable evaluations regarding airports' food/beverage concessions and Wi-Fi services, but not their retail options, have a relevant moderating effect on that relationship. Furthermore, dissatisfaction arising from passengers' interaction with the airline can have negative spillover effects on their satisfaction with the airport. We also found evidence of blame-attribution behavior, in which only delays of internal origin, such as failures in flight management, are significant, indicating that passengers overlook weather-related flight delays. Finally, the results suggest that an empirical specification that does not consider the latent psychosituational traits of passengers produces a relevant overestimation of the absolute effect of flight delays on passenger satisfaction.